SED expression < > meaning-CodePudding

what does the < and > at the beginning and at the end of this sed expression mean?

'/<copyright>/d'

Thanks!

CodePudding user response：

that's the string who's sed searching a match. Everything between the separators, in this case '/', is the string who is searched by sed command.

CodePudding user response：

The d command removes all lines matching the preceding regular expression. In this concrete case, it removes all lines containing <copyright>, no matter where it appears. This is not like the s replace command where only what was matched is replaced, here the whole line is affected. For example:

$ cat input.txt
1. This is a copyright-protected document.
2. The <copyright> is inside some sort of tag.
3. The "copyright" here is quoted.
4. The copyright is reserved.
5. XcopyrightY is not a word.

$ sed '/<copyright>/d' < input.txt
1. This is a copyright-protected document.
3. The "copyright" here is quoted.
4. The copyright is reserved.
5. XcopyrightY is not a word.

As you can see, only line 2 was removed, because it is the only line matching the search expression (regex).

By the way, literal < and > characters are not to be mistaken for the regex extensions \< and \> which match word boundaries. For example:

$ sed '/\<copyright\>/d' < input.txt
5. XcopyrightY is not a word.

$ sed 's/\<copyright\>/(c)/g' < input.txt
1. This is a (c)-protected document.
2. The <(c)> is inside some sort of tag.
3. The "(c)" here is quoted.
4. The (c) is reserved.
5. XcopyrightY is not a word.